Burrow Plasticity in the Deep-Sea Isopod Bathynomus

View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Shinshu University Institutional Repository

Burrow Plasticity in the Deep-Sea Isopod Bathynomus doederleini (Crustacea: Isopoda: Cirolanidae) Author(s) :Toshinori Matsui, Tohru Moriyama and Ryuichi Kato Source: Zoological Science, 28(12):863-868. 2011. Published By: Zoological Society of Japan DOI: http://dx.doi.org/10.2108/zsj.28.863 URL: http://www.bioone.org/doi/full/10.2108/zsj.28.863

BioOne (www.bioone.org) is a nonprofit, online aggregation of core research in the biological, ecological, and environmental sciences. BioOne provides a sustainable online platform for over 170 journals and books published by nonprofit societies, associations, museums, institutions, and presses. Your use of this PDF, the BioOne Web site, and all posted and associated content indicates your acceptance of BioOne’s Terms of Use, available at www.bioone.org/page/terms_of_use. Usage of BioOne content is strictly limited to personal, educational, and non-commercial use. Commercial inquiries or rights and permissions requests should be directed to the individual publisher as copyright holder.

BioOne sees sustainable scholarly publishing as an inherently collaborative enterprise connecting authors, nonprofit publishers, academic institutions, research libraries, and research funders in the common goal of maximizing access to critical research. ZOOLOGICAL SCIENCE 28: 863–868 (2011) ¤ 2011 Zoological Society of Japan

Burrow Plasticity in the Deep-Sea Isopod Bathynomus doederleini (Crustacea: Isopoda: Cirolanidae)

Toshinori Matsui1, Tohru Moriyama2* and Ryuichi Kato3

1Faculty of Textile Science and Technology, Shinshu University, 3-15-1 Tokida, Ueda, Nagano 386-8567, Japan 2Young Researchers Empowerment Center, Shinshu University, 3-15-1 Tokida, Ueda, Nagano 386-8567, Japan 3Graduate School of Science and Technology, Shinshu University, 3-15-1 Tokida, Ueda, Nagano 386-8567, Japan

We investigated whether the deep-sea isopod Bathynomus doederleini has the capacity to change burrow length in response to changes in environmental conditions. We observed burrowing behavior in individuals that were placed on substrates with either simple (ST) or complex (CT) surface topographies. Individuals in the ST group (N = 10) constructed seven burrows. The mean ratio of the burrow length to body length was 1.8. The individuals in the CT group (N = 10) constructed eight burrows with a mean ratio of burrow length to body length of 2.5. Thus the burrows were significantly longer in the CT group. In addition, the isopods in the CT group often incorporated a chamber in the mid-section of the burrow. Our results may be used to infer the determinants of burrow morphology and speculate about the lifestyle of this species in the deep sea.

Key words: burrow, plasticity, deep sea, isopod, Crustacea, Bathynomus doederleini

was a shelter, not a nest. More recently, Iwasaki et al. INTRODUCTION (2001) reported that burrow construction was initiated when- The mesopelagic zone (200–1,000 m below sea level) ever an individual made contact with a vertical acrylic wall, contains the greatest biomass and diversity of animal life in which was placed away from the lateral side of the aquarium the ocean (Warrant and Locket, 2004). The isopod on a gelatin substrate. In the absence of this wall, the indi- Bathynomus is distributed primarily within this zone, and is vidual dug along the side of the aquarium. These observa- classified as part of the littoral-bathyal benthos (Wolf, 1970). tions suggest that the burrowing behavior of B. doederleini This isopod is a highly mobile, voracious scavenger that is induced by mechanical stimulation of the head. Further- dominates the fauna associated with large dead animals more, these two studies suggest that the length of the bur- (Sekiguchi et al., 1981). row is equal to that of the body in both natural and artificial There has been an increasing awareness that highly substrates. mobile epibenthic invertebrates play a significant role in the In contrast, Matsuoka and Koike (1980) reported that dynamics of deep-sea benthic communities (Hessler et al., the burrow length of a Miocene period Bathynomus sp. was 1978). Thus, it is important to understand the biology of approximately six times longer than that of the occupant’s Bathynomus and its role in the deep benthic environment body length. In addition, the burrow had an expanded (Tsuo and Mok, 1991). Researchers have studied the clas- region, approximately twice as wide as the occupant’s body sification, geographical distribution, and development of width, in the mid-section of the burrow. Thus, it has Bathynomus (Holthuis and Mikulka, 1972; Sekiguchi et al., remained unclear whether there are differences in burrow 1982; Tsuo and Mok, 1991), but little is known about the length among captive and wild B. doederleini. ecology and behavior of this species because of the diffi- To address this, we attempted to duplicate features of culty observing animals in situ at such great depths. How- the B. doederleini natural habitat in the aquarium and eval- ever, Bathynomus is able to survive for several months in uate whether the morphology of burrows dug by extant B. aquaria at atmospheric pressure, affording researchers the doederleini resembled those dug in the Miocene. The opportunity to study its behavior. For example, Sekiguchi surface topography of the sea floor is highly complex, in part (1985) reported burrowing behavior in a captive Bathynomus due to the presence of carcasses, e.g., whale skeletons doederleini individual. Because the burrow was equal in (Gage & Tyler, 1991). Thus, we hypothesized that B. length to the isopod, the author concluded that the burrow doederleini construct longer burrows in topographically complex substrates when compared with simple substrates, * Corresponding author. Phone: +81-(0)268-21-5589; which are typically used in experiments with captive B. Fax : +81-(0)268-21-5884; doederleini. To test this, we documented burrowing behav- E-mail: [email protected] ior in B. doederleini that were placed on substrates with sim- doi:10.2108/zsj.28.863 ple or complex surface topography. 864 T. Matsui et al.

of 100 cm from the surface of the substrate. The camera was con- MATERIALS AND METHODS nected to a digital high-resolution recorder (Sharp DV-ACW72). The Animals LEDs were turned on during the experiments to allow recording We collected specimens of B. doederleini using baited traps in while the aquaria were in the dark. Sagami Bay, Japan. The animals were then transferred to tanks We randomly selected 20 isopod individuals from the holding containing recirculated sea water (10 ± 0.2°C; pH 8 ± 0.2; 30 ± 1 tanks and divided them into two groups: ST (N = 10) and CT (N = ‰S) and held in the dark for six months. Each individual was fed 10). We measured the body length and width of each individual and slices of raw squid (5 g) once every two weeks. We placed the indi- determined their sex (Tsuo and Mok, 1991). Each isopod was individuals in separate buckets while feeding. vidually placed on the center of the aquarium on the surface of the substrate (ST group) or in one of the tubes (CT group). We Experimental aquaria recorded the behavior of animals in both groups during a 6 h period. We used a transparent cylindrical aquarium (60 cm in diameter, At the end of the experiment, we measured the burrow length. 40 cm in height) that was placed in the laboratory (room tempera- If an individual was found in the burrow, it was gently removed by ture, 15°C). We created two types of substrate having differing sur- hand to avoid damage to the burrow. We injected green fluid into face topography, simple (ST) and complex (CT). For both groups, each empty burrow to create a cast. The burrow morphology was the aquaria were filled with 0.3% agar to a depth of 25 cm. The con- photographed using a digital camera (Nikon Coolpix P5100). centration of agar was determined during preliminary tests in which Each individual was used only once. The agar substrate and B. doederleini individuals were placed in aquaria filled with various seawater were changed between individuals. The plastic tubes concentrations agar substrata. We observed the behavior of the iso- were also washed with water between individuals. pods to ensure they were able to crawl and burrow in the agar. After the agar set, sea water was poured into the aquarium to a depth of Statistical analyses 10 cm above the surface of the agar substrate. The experimental We evaluated the differences in mean body length, mean body setup was left until the temperature at the center of the substrate width, mean time from the initial movement on the substrate to the reached 16 ± 1°C. completion of a burrow, and the mean time spent in a burrow In half of the aquaria, we placed five plastic, transparent T- between the two groups using an unpaired Student’s t-test. This test shaped tubes (6 cm diameter, 12 cm trunk length, 9 cm branch was also used to evaluate differences in the mean ratio of the bur- length) on the surface of the substrate (Fig. 1). The tubes were con- row length to the occupant’s body length between burrows with and nected to the base of the aquarium using thin plastic poles to pre- without chambers in the CT group. We evaluated the difference in vent movement. The distance between each tube and the aquarium the mean ratio of burrow length to the occupant’s body length wall or a neighboring tube was sufficient to allow passage of the iso- between the treatment groups using Welch’s t-test. We tested for pods. This more complex, maze-like substrate surface was intended normality of the data using a Shapiro-Wilk test. If the data were nor- to duplicate the surface topography of the animals’ natural habitat. mally distributed, we used an F-test to evaluate differences in the Individuals were able to move freely and were expected to occa- standard deviation between two samples. We used Fisher’s exact sionally enter the T-tubes, resulting in mechanical stimulation of probability test to evaluate differences in the sex ratio and in the their body surface. The animals are also forced to select a direction incidence of specimens that constructed burrows between the treat- at the junction of the T. We believe this behavioral pattern simulates ment groups. We used Spearman’s rank correlation test to evaluate activity in their natural habitat, as they are likely to encounter car- the direction and strength of the relationship between total tube- casses and animal skeletons that resemble a maze-like structure. passing time and the ratio of burrow length to the occupant’s body length. We used the Mann-Whitney U test to evaluate differences Experiments in the mean time spent in the tubes between specimens that con- We installed a CCD camera with built-in infrared LEDs structed burrows and those that did not in the CT group. (Corona-Dengyo CZ-100, Japan) above the aquarium at a distance Differences with a P value of < 0.05 were considered significant. The Shapiro-Wilk test was performed using Origin 8.5 (OriginLab Corporation, Northampton, MA, USA). All other tests were conducted using Ekuseru-toukei 2006 (Social Survey Research Information Co., Ltd., Tokyo, Japan). The data are presented as the mean ± S.E.

RESULTS Body length, body width, and sex ratio The mean body length, mean body width, and the sex ratio (male:female) were 9.6 ± 0.4 cm, 4.1 ± 0.2 cm, and 5:5, respectively (ST, N = 10) and 10.0 ± 0.7 cm, 4.3 ± 0.3 cm, and 6:4 (CT, N = 10). There was no difference among the treatment groups (body length: F9,9 = 2.33, P = 0.223; t = 0.983, df = 18, P = 0.1692; body width: F9,9 = 2.53, P = 0.184; t = 1.175, df = 18, P = 0.128; sex ratio: P = 0.500). All individuals were maturing (Soong and Mok, 1994), but none of the females had eggs.

Burrow construction behavior Fig. 1. Experimental aquarium illustrating the setup of the complex The position of each individual was plotted each second substrate topography (CT). The tubes were absent in the simple sur- (Fig. 2) to document the steps involved in burrow construc- face topography (ST) setup. Seawater was added before the ani- tion. Based on our observations, the process was divided mals were introduced. into three stages: Burrow Plasticity in B. doederleini 865

eral minutes (mean: 4.4 ± 1.7 min, N = 10) when first placed into the aquarium. After this initial period, all the animals uncurled and began moving (walking or swim- ming) and excavating depressions in the agar along the wall of the aquarium (Fig. 3A). We did not observe any depressions being excavated away from the wall. The isopods dug into the agar using their thoracic legs and removed the debris using water currents produced by beating their swim- merets. The depressions were mortar- shaped structures each approximately a quarter of the body length of the individual (Fig. 3A). Several individuals excavated multiple depressions, forming a groove along the wall (Fig. 3B). The isopods in the CT group were often observed moving through the tubes, entering head-first in most cases, but backwards in some. We also observed animals somersaulting in the tubes in a limited number of instances.

(b) Burrow construction Following surface excavation, several individuals spent time repeatedly digging in one or two specific depressions located along the wall. As they dug, their body was positioned at around a 30° inclination, head down (Fig. 3C). All individuals exited the depression by moving backwards. We defined a depression as a ‘burrow’ when its length was approximately equal to the body length of the individual digging the burrow (Fig. 3C, D). The mean time from the initial motion on the substrate to Fig. 2. A typical time sequence showing changes in the behavior of a specimen in the CT the completion of a burrow (or the group. The horizontal axis represents time (30 min units). The vertical axis represents positions first burrow in the case of a speci- in the aquarium. T, S, D and B are illustrated in the inset figure. T: tubes; S: surface of substrate; men that constructed two burrows) D: depressions; B: a burrow. Each black horizontal line with a closed circle at the left end and was 126.8 ± 89.4 min (N = 6) and open circle at the right end represents continuous occupation at a specific position. The closed 96.0 ± 66.9 min (N = 7) in the ST circle denotes the initial time and the open circle represent the ending time. A thick black arrow and CT groups, respectively (F5,6 = denotes the time when a depression became a burrow (3:09:15). The specimen crept out of the 1.39, P = 0.694; t = 0.706, df = 11, burrow headfirst (head-first exit) at 4:01:15, 4:02:30, and 5:38:50. The individual entered the bur- P = 0.247). row ‘telson-first’ (telson-entry) at 4:02:23. We observed seven burrows in the ST group. Six individuals (a) Surface excavation dug a single burrow, and one individual dug two. The four The individuals in the ST group conglobated and remaining individuals only dug depressions. Similarly, we remained balled up for a short period (mean: 35.5 ± 11.2 s, observed eight burrows in the CT group. Seven individuals N = 10) when first placed on the agar substrate. Conversely, dug a single burrow and one individual dug two. The remain- the isopods in the CT group remained conglobated for sev- ing three individuals only dug shallow depressions. 866 T. Matsui et al.

Specimens that constructed burrows were termed ‘burrowers’. Those that continued to move about the substrate without constructing a burrow were termed ‘wanderers’. There was no significant difference in the incidence of burrowers between the treatment groups (P = 0.500).

(c) Burrow modification phase After completion of the burrow, the occupants spent the majority of time in their burrow, either digging or resting. Four of seven burrows were not elongated any further in the ST group (Fig. 3D). These four burrows were only slightly longer than the occupying individuals’ body length. In contrast, all eight burrows were elongated to almost twice the occupants’ body length in the CT group (Fig. 3E). Four burrows were even elongated to nearly three times the occupants’ body length and curved at the base of the aquarium. In these long burrows, the occupants sometimes somersaulted in the mid-section of the burrow, moved upward, then paused at the entrance before crawling out head-first. The somersaulting motion compacted the wall of the burrow, resulting in the formation of a chamber (Fig. 4).

Fig. 3. Depressions and burrows. Typical depressions (A). Multiple depressions formed a Burrow length groove (B). A typical burrow whose length was almost equal to that of the specimen (C) and into We used the ratio of the bur- which green fluid agar was casted (D). A typical burrow whose length was almost twice that of row length to the occupant’s body the specimen and into which green fluid agar was casted (E). length as an index for the length of a burrow (Rb/o). The mean Rb/o was significantly higher in CT (N = 8) than ST (N = 7) (F6,7 = 5.66, P = 0.039; t (Welch) = 2.22, df = 7.84, P = 0.029, Fig. 5). Within CT, the mean Rb/o for burrows with chambers (N =

Fig. 4. Profile view of a chamber in the mid-section (surrounded by dashed circle) of a burrow in the CT group. Fig. 5. Mean Rb/o for the CT and ST groups (*P < 0.05). Burrow Plasticity in B. doederleini 867

4) was significantly higher than for those without them (N = each individual in the CT group. Among the burrowers, there 4) (F3,3 = 1.26, P = 0.855; t = 5.04, df = 6, P = 0.0023, Fig. was no significant correlation between the time to transit 6). through the tubes and Rb/o (N = 7, rs = 0.429, P = 0.337). In one instance, a single individual constructed two burrows. Time in burrow Thus, the two Rb/o values were summed and the total value We calculated the total time spent in a burrow for each was used for analysis. The mean time for wanderers (N = burrower in the ST and CT groups. The mean time for the 50) was significantly longer than for burrowers (N = 47) (U = CT group (N = 8) was significantly longer than for the ST 825.5, z = 2.52, P = 0.012) (Fig. 8). group (N = 7) (F = 1.02, P = 0.994; t = 1.89, df = 6, P = 6,7 DISCUSSION 0.040) (Fig. 7). We demonstrated that substrate surface topography is Time in tubes correlated with the length of burrows constructed by B. We calculated the time spent in the plastic tubes for doederleini. More complex substrate surface topography was associated with construction of longer burrows (Fig. 5). In addition, half of these longer burrows (4 of 8) had chambers in the mid-section (Fig. 6), suggesting they were used as a nest. The presence of these two features is consistent with observations of the fossilized burrows of Miocene period Bathynomus sp. These fossilized burrows were approximately six times longer than the constructing individual’s body length, and contained an expanded region that was approximately twice as wide as the occupying individual (Matsuoka and Koike, 1980). Our results suggest that extant B. doederleini construct a burrow that is approximately 2.5 times its body length and incorporates a chamber in the mid-section of the burrow. The chamber is constructed by initiating a vigorous somersaulting motion. Once construction was complete, the occu- Fig. 6. Mean Rb/o for ‘no chamber’ and ‘chamber’ groups (**P < pants were sometimes observed somersaulting, moving 0.01). upward, pausing at the opening for a few minutes, then crawling out of the burrow head-first. The pause at the burrow entrance may be indicative of a ‘sit and wait’ foraging strategy, such as that observed in the deep-sea isopod, Natatolana borealis (Johansen and Brattegard, 1998). Natatolana borealis is also a voracious scavenger that inhabits soft substrates offshore (Wong and Moore, 1995). This species (body length: 2 cm) also constructs a perma- nent U-shaped burrow that extends 7–10 cm into the substrate (Taylor and Moore, 1995). Natatolana borealis enters one of the openings head-first then moves through the burrow to the opposite entrance/exit. Taylor and Moore (1995) observed individuals lying stationary near the entrance with only the anterior section visible and the second antennae pointing upwards above the sediment. These individuals Fig. 7. Mean time spent in a burrow for the CT and ST groups emerged intermittently to feed when carcass odors were (*P < 0.05). detected in the water. Only one of the seven burrows constructed on the simple substrate incorporated a chamber. The length of the remaining burrows was slightly longer than the length of the occupants (Mean Rb/o = 1.45 ± 0.22). These individuals were never observed somersaulting in the burrows, and exited by crawling backwards. Given that this animal is a highly mobile, voracious scavenger, we hypothesize that it digs two types of burrows. The first is a short burrow that functions as a temporary shelter during foraging and the second is a deep burrow that the animal uses as a nest. Burrowing reduces predation risk by decreasing accessibility to predators (Zwarts et al., 1996) and deeper burrows may provide a better refuge against Fig. 8. Mean time to transit through the tubes for burrowers and predators (Zwarts and Wanink, 1993). The depth of a bur- wanderers (*P < 0.05). row represents a tradeoff between energy expenditure and 868 T. Matsui et al. benefit. Thus, B. doederleini occupied deeper burrows for Science and Technology Agency Special Coordination Funds for longer periods than a shallow burrow (Fig. 7). Promoting Science and Technology 2007. Sekiguchi (1985) observed burrow construction in B. REFERENCES doederleini and noted that the burrows were approximately equal in length to the body of the occupant. We also noted Gage DG, Tyler PA (1991) Deep-Sea Biology: A Natural History of that the individuals moved rapidly, head-first into the bur- Organisms at the Deep-Sea Floor. Cambridge University Press, rows when disturbed, then later exited backwards, and con- UK Hessler RR, Ingram, CL, Yayanos AA, Burnett BR (1978) Scaveng- cluded that the burrow most likely represents a temporary ing amphipods from the floor of the Philippine Trench. Deep- shelter, not a nest. Taken together, these observations sug- Sea Res 25: 1029–1047 gest that B. doederleini individuals perceive a greater risk of Holthuis LB, Mikulka WR (1972) Notes on the deep-sea isopods of predation in substrates with simple topography, so are com- the genus Bathynomus A. Milne Edwards, 1879. Bull Mar Sci pelled to dig a burrow that functions as a temporary shelter. 22: 575–591 Although the individuals on more complex substrates were Iwasaki M, Ohata A, Okada Y, Sekiguchi H, Niida A (2001) Func- clearly motivated to burrow deeply, the factors for this moti- tional organization of anterior thoratic stretch receptors in the vation remain unclear. The ability of a predator to locate and deep-sea isopod Bathynomus doederleini: behavioral, morpho- access subsurface prey is likely related to the substrate sur- logical and physiological studies. J Exp Biol 204: 3411–3423 face complexity. More complex substrates likely increase Johansen PO, Battegard T (1998) Observation on behavior and distribution of Natatolana borealis (Lilljeborg) (Crustacea, the difficulty of locating prey and/or accessing burrows. Isopoda). Sarsia 83: 347–360 There is a need for further experiments to evaluate whether Matsuoka K, Koide K (1980) Bathynomus (Crustacea: Isopoda) this feature of complex substrates motivates B. doederleini from the Miocene Yatsuo Formation, Toyama Prefecture, Cen- individuals to burrow more deeply. tral Japan Bull Mizunami Fossil Mus 7: 51–58 We found no difference in the time from initial movement Richardson, H (1909) Isopods collected in the northwest pacific by on the substrate to the completion of a burrow among the the U S bureau of fisheries steamer “albatross” in 1906. Proc U treatment groups. Similarly, there was no difference in the S Nat Mus 37: 75–129 number of burrows constructed between the two groups. Richardson, H (1910) Marine isopods collected in the Philippines by Taken together, these observations suggest that both the U S fisheries steamer Albatross in 1907–8. Docum U S Bur spontaneous and passive tube-passing have no effect on an Fish 736: 1–44 Sekiguchi H (1985) Note on the burrow of a giant deep-sea isopod individual’s motivation for burrowing. However, other envi- Bathynomus doederleini (Flabelifera: Cirolanidae). Proc Lap ronmental cues (e.g., seafloor topography) likely have an Soc Syst Zool 31: 26–29 effect on the motivation to burrow. Sekiguchi H, Yamaguchi Y, Kobayashi H (1981) Bathynomus We observed “wandering” individuals in both treatment (Isopoda: Cirolanidae) attacking sharks caught in a gill-net. Bull groups. These individuals dug several shallow depressions Fac Fish Mie Univ 8: 11–17 in the agar, but never completed a burrow. In the simple Sekiguchi H, Yamaguchi Y, Kobayashi H (1982) Geographical distri- substrate, these individuals repeatedly moved about the sur- bution of scavenging giant isopods Bathynomids in the North- face and excavated depressions on the substrate. However, western Pacific. Bull Japan Soc Sci Fish 48: 499–504 on the more complex surface, these individuals often Soong K, Mok H-K (1994) Size and maturity stage observations of passed through the tubes, entering headfirst in most cases the deep-sea isopod Bathynomus doederleini Ortomann, 1984 (Flabellifera: Cirolanidae), in eastern Taiwan. J Crust Biol 14: but backwards in others. The mean time to pass through the 72–79 tubes was almost twice as long in the wanderers as in the Taylor AC, Moore PG (1995) The burrows and physiological adapta- burrowers (Fig. 8), suggesting that the wanderers used the tions to a burrowing lifestyle of Natatolana boreali (Isopoda: tubes for temporary shelter. Cirolanidae). Mar Biol 123: 805–814 In summary, we demonstrated phenotypic plasticity in Tso S-F, Mok H-K (1991) Development, reproduction and nutrition the burrows of the deep-sea isopod B. doederleini in of the giant isopod Bathynomus doederleini Ortmann, 1894 response to sea floor topography complexity. The capacity (Isopoda, Flabellifera, Cirolanidae). Crustaceana 61: 141–154 to alter burrow morphology in response to environmental dif- Warrant EJ, Locket NA (2004) Vision in the deep sea. Biol Rev 79: ferences suggests they have a highly organized central ner- 671–712 vous system that is able to assess their environment. To Wong YM, Moore PG (1995) Biology of feeding in the scavenging isopod Natatolana borealis (Isopoda: Cirolanidae). Ophelia 43: identify the neurophysiological pathways involved in making 181–196 this decision, we intend to investigate which stimuli the ani- Zwarts L, Wanink JH (1993) How the food supply harvestable by mals are responsive to. wanders in the Wadden Sea depends on the variation in energy ACKNOWLEDGMENTS density, body weight, biomass, burying depth and behavior of tidal –flat invertebrates. Netherlands J Sea Res 31: 441–476 We are grateful to Dr. Makoto Kurokawa and his laboratory Zwarts L, Ens BJ, Gross-Custard JD, Hulscher JB, Le V Dit Durell staff, as well as Dr. Kosuke Tanaka for support throughout this SEA (1996) Causes of variation in prey profitability and its con- study. We thank Mr. Junichi Suzuki for collection of the specimens sequences for the intake rate of the Oystercatcher Haematopus and Ina Food Industry Co., Ltd., for assistance making the agar ostralegus. Ardea 84A: 229–268 substrate. We also thank Dr Yoshiyuki Matsumura for input into this manuscript. This research was supported in part by the Japan (Received March 11, 2011 / Accepted June 27, 2011)